基于弹性波正演和深度学习数据拓展的全波形反演方法研究

全波形反演是一种利用地震波波形信息分析地下介质速度分布的高分辨率速度反演方法。该方法被广泛应用于油气资源勘探、大尺度地球壳幔结构分析、防灾减灾和微地震监测等众多领域。可靠的低频观测数据和稳定有效的计算方法是全波形反演能获得高分辨率反演结果的重要条件。目前，虽然全波形反演方法在部分质量较好且远偏移距信息充足的海底OBC/OBS数据上取得了一系列令人振奋的反演效果，但对于更为一般的，远偏移距信息不足且缺失低频的陆地等方式获得的观测数据，全波形反演方法的效果仍然存在较大的提升空间。为此，本文主要针对方法目前所面临的弹性效应、周波跳跃和数据增强等问题进行了深入的研究。

全波形反演方法的计算成本较高，目前在实际应用中通常基于声波假设近似处理实际的弹性观测数据，这在减少方法计算成本的同时也引入了一定的弹性效应误差。为缓解弹性效应对反演结果的不利影响，本文在声波全波形反演方法的基础上提出了一种弹性-声波混合反演方案。在方法的正演模拟阶段，使用弹性波方程替换了原本的声波方程，用于更真实地模拟地震波场的传播，以降低弹性效应所导致的波形扰动误差；对于受弹性效应影响更小的反演阶段，本文仍然基于声波方程计算波形扰动所对应的模型梯度，保证了方法整体的计算成本在实际的可接受范围内。陆地实测数据的测试结果表明，混合反演方法提升了反演结果的清晰度，但弹性效应并不是影响方法收敛的主要因素，因此本文对方法进行更进一步的改进。

周波跳跃是影响全波形反演效果的主要问题之一，近年来采用机器学习方法人为构建低频信息的方案，为快速有效地解决这一问题提供了新的思路。本文选择采用生成式人工智能方法中的循环生成对抗神经网络（CycleGAN）构建实测数据中缺失的低频信息，从而缓解周波跳跃问题。与一般的局限与合成数据训练的监督学习类方法不同，CycleGAN属于无监督学习类方法，可以将实测数据纳入到网络的训练过程中，更好地保证了方法在实际应用中的有效性。此外，为了克服无监督学习类方法稳定性较弱的缺点，本文还设计了一种复合形式的目标数据集和额外的损失函数，以进一步提升网络输出结果的准确性。通过定量的数值测试，验证了改进方案的有效性。

实测数据中存在的噪声，坏道和衰减等干扰因素同样影响着方法的反演效果，为此本文的研究进一步发掘了CycleGAN网络对地震数据的增强功能。CycleGAN具有自监督学习能力，能够同时学习数据的多种特征信息，并直接建立输入数据至目标数据集之间的映射关系。当目标数据被设置为包含低频的无干扰数据时，训练好的网络可以同时具备低频构建、去噪、数据补全和能量补偿等多种功能，简化了实测数据的预处理流程。本文通过数值模拟对方法的相应功能进行了详细的测试，并给出了定量的评价结果。

最后，本文在含有测井速度资料的实测数据上对方法进行了测试，并从多个角度论证了方法的可行性和有效性。测试结果表明本文的相关研究有助于提升全波形反演方法对实测数据的反演效果，使方法在应用于一般的低质量无低频观测数据时更为有效，有利于方法在更多实际场景中的应用和推广。此外，本文中与深度学习方法相关的程序已进行了开源共享，以促进相关研究更好地发展。

[

[1] AKI K, RICHARDS P. Quantitative seismology[M]. W.H. Freeman and Company, 1980.

[2] SCHUSTER G. Seismic inversion[M]. Society of Exploration Geophysicists, 2017.

[3] VIRIEUX J, OPERTO S. An overview of full-waveform inversion in exploration geophysics[J]. Geophysics, 2009, 74(6): WCC1-WCC26.

[4] YAO G, WU D, WANG S X. A review on reflection-waveform inversion[J]. Petroleum Science,2020, 17: 334-351.

[5] GOODFELLOW I, BENGIO Y, COURVILLE A. Deep learning[M]. MIT Press, 2016.

[6] LECUN Y, BENGIO Y, HINTON G. Deep learning[J]. Nature, 2015, 521: 436-444.

[7] YU S, MA J. Deep learning for geophysics: current and future trends[J]. Reviews of Geophysics,2021, 59(3): e2021RG000742.

[8] MOUSAVI S M, BEROZA G C. Deep-learning seismology[J]. Science, 2022, 377(6607):eabm4470.

[9] LAILLY P. The seismic inverse problem as a sequence of before stack migrations[J]. Conferenceon Inverse Scattering—Theory and Application, SIAM, Philadelphia, 1983: 206-220.

[10] TARANTOLA A. Inversion of seismic reflection data in the acoustic approximation[J]. Geophysics,1984, 49(8): 1259-1266.

[11] GAUTHIER O, VIRIEUX J, TARANTOLA A. Two‐dimensional nonlinear inversion of seismicwaveforms: numerical results[J]. Geophysics, 1986, 51(7): 1387-1403.

[12] PICA A, DIET J, TARANTOLA A. Nonlinear inversion of seismic reflection data in a laterallyinvariant medium[J]. Geophysics, 1990, 55: 284-292.

[13] CRASE E. Robust elastic nonlinear waveform inversion: application to real data[J]. Geophysics,1990, 55: 527-538.

[14] TARANTOLA A. A strategy for nonlinear elastic inversion of seismic reflection data[J]. Geophysics,1986, 51(10): 1893-1903.

[15] MORA P. Nonlinear elastic inversion of multioffset seismic data[J]. Geophysics, 1986, 52:1211-1228.

[16] MORA P. Elastic wave-field inversion of reflection and transmission data[J]. Geophysics,1988, 53: 750-759.

[17] PRATT R, WORTHINGTON M. Inverse theory applied to multi-source cross-hole tomography,part 1: acoustic wave-equation method[J]. Geophysical Prospecting, 1990, 38: 287-310.

[18] PRATT G, SHIN C, HICKS. Gauss–Newton and full Newton methods in frequency–spaceseismic waveform inversion[J]. Geophysical Journal International, 1998, 133(2): 341-362.

[19] PRATT R G. Seismic waveform inversion in the frequency domain, part 1: theory and verificationin a physical scale model[J]. Geophysics, 1999, 64(3): 888-901.

[20] LUO Y, SCHUSTER G T. Wave-equation traveltime inversion[J]. Geophysics, 1991, 56(5):645-653.

[21] SHIN C, CHA Y. Waveform inversion in the Laplace domain[J]. Geophysical Journal International,2008, 173: 922-931.

[22] PYUN S, SHIN C, LEE H, et al. 3D elastic full waveform inversion in the Laplace domain[J].SEG Technical Program Expanded Abstracts, 2008: 1976-1980.

[23] CHA Y H, SHIN C. 2D Laplace‐domain waveform inversion using adaptive finite elementmethod[J]. SEG Technical Program Expanded Abstracts, 2009: 2263-2267.

[24] YANG Q, MALCOLM A. Frequency domain full-waveform inversion in a fluid-saturatedporoelastic medium[J]. Geophysical Journal International, 2020, 225(1): 68-84.

[25] TROMP J, TAPE C, LIU Q. Seismic tomography, adjoint methods, time reversal and bananadoughnutkernels[J]. Geophysical Journal International, 2005, 160(1): 195-216.

[26] PLESSIX R E. A review of the adjoint-state method for computing the gradient of a functionalwith geophysical applications[J]. Geophysical Journal International, 2006, 167: 495-503.

[27] ZHAO L, JORDAN T H, OLSEN K B, et al. Fréchet kernels for imaging regional earth structurebased on three-dimensional reference models[J]. Bulletin of the Seismological Society ofAmerica, 2005, 95(6): 2066-2080.

[28] ZHAO L, CHEN P, JORDAN T H. Strain green’s tensors, reciprocity, and their applicationsto seismic source and structure studies[J]. Bulletin of the Seismological Society of America,2006, 96(5): 1753-1763.

[29] CHEN P, JORDAN T H, ZHAO L. Full three-dimensional tomography: a comparison betweenthe scattering-integral and adjoint-wavefield methods[J]. Geophysical Journal International,2007, 170(1): 175-181.

[30] SIRGUE L, I. BARKVED O, P. VAN GESTEL J, et al. 3D waveform inversion on valhallwide-azimuth OBC[J]. European Association of Geoscientists & Engineers, 2009: 2214-4609.

[31] 胡光辉, 王立歆, 方伍宝. 全波形反演方法及应用[M]. 石油工业出版社, 2014.

[32] 祝贺君, 刘沁雅, 杨继东. 地震学全波形反演进展[J]. 地球与行星物理论评(中英文), 2023,54: 287-317.

[33] WARNER M, RATCLIFFE A, NANGOO T, et al. Anisotropic 3D full-waveform inversion[J].Geophysics, 2013, 78(2): R59-R80.

[34] MEI J, TONG Q. A practical acoustic full waveform inversion workflow applied to a 3D landdynamite survey[J]. SEG Technical Program Expanded Abstracts, 2015: 1220-1224.

[35] DELLINGER J, ROSS A, MEAUX D, et al. Wolfspar®, an“FWI-friendly”ultralow-frequencymarine seismic source[J]. SEG Technical Program Expanded Abstracts, 2016: 4891-4895.

[36] CHI B, DONG L, LIU Y. Full waveform inversion method using envelope objective functionwithout low frequency data[J]. Journal of Applied Geophysics, 2014, 109: 36-46.

[37] LI Y, DEMANET L. Extrapolated full-waveform inversion: an image-space approach[J]. SEGTechnical Program Expanded Abstracts, 2017: 1682-1686.

[38] CHEN W, JIAL X, HU C. Time-domain multiscale FWI based on a time-shift method and lowfrequencydata reconstruction[J]. SEG Technical Program Expanded Abstracts, 2019: 1570-1574.

[39] CHIU S K. Low-frequency extrapolation using higher-frequency multi-channel prediction filters[J]. SEG Technical Program Expanded Abstracts, 2020: 2835-2839.

[40] OVCHARENKO O, KAZEI V, KALITA M, et al. Deep learning for low-frequency extrapolationfrom multioffset seismic data[J]. Geophysics, 2019, 84(6): R989-R1001.

[41] SUN H, DEMANET L. Extrapolated full-waveform inversion with deep learning[J]. Geophysics,2020, 85(3): R275-R288.

[42] FANG J, ZHOU H, ELITA LI Y, et al. Data-driven low-frequency signal recovery using deeplearningpredictions in full-waveform inversion[J]. Geophysics, 2020, 85(6): A37-A43.

[43] WANG M, XU S, ZHOU H. Self-supervised learning for low frequency extension of seismicdata[J]. SEG Technical Program Expanded Abstracts, 2020: 1501-1505.

[44] OVCHARENKO O, KAZEI V, ALKHALIFAH T A, et al. Multi-task learning for lowfrequencyextrapolation and elastic model building from seismic data[J]. IEEE Transactionson Geoscience and Remote Sensing, 2022, 60: 1-17.

[45] WANG Z, LIU G, DU J, et al. Low-frequency extrapolation of prestack viscoacoustic seismicdata based on dense convolutional network[J]. IEEE Transactions on Geoscience and RemoteSensing, 2022, 60: 1-13.

[46] GU Z, CHAI X, YANG T. Deep-learning-based low-frequency reconstruction in full-waveforminversion[J]. Remote Sensing, 2023, 15(5): 1387.

[47] SUN H, SUN Y, NAMMOUR R, et al. Learning with real data without real labels: a strategyfor extrapolated full-waveform inversion with field data[J]. Geophysical Journal International,2023, 235(2): 1761-1777.

[48] 曹伟. 基于深度学习的地震速度反演研究[D]. 东北石油大学, 2023.

[49] 李媛媛. 多级优化的时间域波形反演方法研究[D]. 中国石油大学(华东), 2020.

[50] 杜泽源. 多尺度地震全波形反演方法研究[D]. 中国石油大学(华东), 2022.

[51] 董士琦. 基于波场局部尺度匹配的时间域全波形反演研究[D]. 吉林大学, 2022.

[52] 李振春, 王自颖, 黄建平, 等. 基于波数域梯度场分解的多尺度波形反演方法[J]. 地球物理学报, 2022, 65: 2693-2703.

[53] 杨瑞冬, 黄建平, 杨振杰, 等. 基于双对角通量校正的多尺度全波形反演[J]. 石油地球物理勘探, 2022, 57: 1120-1128.

[54] BUNKS C, SALECK F M, ZALESKI S, et al. Multiscale seismic waveform inversion[J].Geophysics, 1995, 60(5): 1457-1473.

[55] BOONYASIRIWAT C, VALASEK P, ROUTH P, et al. An efficient multiscale method for timedomainwaveform tomography[J]. Geophysics, 2009, 74(6): WCC59-WCC68.

[56] SHENG J, LEEDS A, BUDDENSIEK M, et al. Early arrival waveform tomography on nearsurfacerefraction data[J]. Geophysics, 2006, 71(4): U47-U57.

[57] SYMES W, CARAZZONE J. Velocity inversion by differential semblance optimization[J].Geophysics, 1991, 56: 654-663.

[58] SYMES W. Migration velocity analysis and waveform inersion[J]. Geophysical Prospecting,2008, 56: 765-790.

[59] WEIBULL W, ARNTSEN B. Automatic velocity analysis with reverse-time migration[J]. Geophysics,2013, 78: S179-S192.

[60] BROSSIER R, OPERTO S, VIRIEUX J. Which data residual norm for robust elastic frequencydomainfull waveform inversion[J]. Geophysics, 2010, 75(3): R37-R46.

[61] ENGQUIST B, FROESE B D. Application of the Wasserstein metric to seismic signals[J].Communications in Mathematical Sciences, 2014, 12(5): 979-988.

[62] MATZEL E, GRAND S P. The anisotropic seismic structure of the East European platform[J].Journal of Geophysical Research: Solid Earth, 2004, 109(B1): 2001JB000623.

[63] CHOI Y, ALKHALIFAH T. Application of multi-source waveform inversion to marine streamerdata using the global correlation norm[J]. Geophysical Prospecting, 2012, 60: 748-758.

[64] MéTIVIER L, BROSSIER R, MéRIGOT Q, et al. Measuring the misfit between seismogramsusing an optimal transport distance: application to full waveform inversion[J]. GeophysicalJournal International, 2016, 205: 332-364.

[65] CHEN J, CHEN Y, WU H, et al. The quadratic Wasserstein metric for earthquake location[J].Journal of Computational Physics, 2018, 373: R43-R62.

[66] 李燕梅, 顾焕申, 许凯, 等. 基于最优输运与循环神经网络的全波形反演[J]. 地球物理学进展, 2022, 37: 2408-2416.

[67] 胡光辉, 贺伟光. 全波形反演中最优化传输函数研究[J]. 石油物探, 2023, 62: 832-849.

[68] GAO Y, TILMANN F, RIETBROCK A. A review of misfit functions for adjoint full waveforminversion in seismology[J]. Geophysical Journal International, 2023, 235(3): 2794-2827.

[69] SEN M, STOFFA P. Nonlinear one-dimensional seismic waveform inversion using simulatedannealing[J]. Geophysics, 1991, 56: 1624-1638.

[70] MOSEGAARD K, TARANTOLA A. Monte Carlo sampling of solutions to inverse problems[J]. Journal of Geophysical Research, 1995, 1001: 12431-12448.

[71] ZHAO Z, SEN M K, DENEL B, et al. A hybrid optimization framework for seismicfull waveform inversion[J]. Journal of Geophysical Research: Solid Earth, 2022, 127(8):e2022JB024483.

[72] GEBRAAD L, BOEHM C, FICHTNER A. Bayesian elastic full-waveform inversion usinghamiltonian monte carlo[J]. Journal of Geophysical Research: Solid Earth, 2020, 125(3):e2019JB018428.

[73] LIU F, ZHANG G, MORTON S A, et al. An effective imaging condition for reverse-timemigration using wavefield decomposition[J]. Geophysics, 2011, 76(1): S29-S39.

[74] XU S, WANG D, CHEN F, et al. Inversion on reflected seismic wave[J]. SEG TechnicalProgram Expanded Abstracts, 2012: 1-7.

[75] KHALIL A, SUN J, ZHANG Y, et al. RTM noise attenuation and image enhancement usingtime-shift gathers[J]. SEG Technical Program Expanded Abstracts, 2013: 3789-3793.

[76] ALKHALIFAH T. Scattering-angle based filtering of the waveform inversion gradients[J].Geophysical Journal International, 2014, 200(1): 363-373.

[77] FEI T W, LUO Y, YANG J, et al. Removing false images in reverse time migration: the conceptof de-primary[J]. Geophysics, 2015, 80(6): S237-S244.

[78] YAO G, DA SILVA N V, WU D. Sensitivity analyses of acoustic impedance inversion withfull-waveform inversion[J]. Journal of Geophysics and Engineering, 2018, 15(2): 461-477.

[79] XU S, ZHANG Y, TANG B. 3D angle gathers from reverse time migration[J]. Geophysics,2011, 76(2): 129-131.

[80] STORK C. Reflection tomography in the postmigrated domain[J]. Geophysics, 2012, 57(5):680-692.

[81] LUO Y, MA Y, WU Y, et al. Full-traveltime inversion[J]. Geophysics, 2016, 81(5): R261-R274.

[82] MUSAYEV K, HACKL K, BAITSCH M. Identification of the velocity field of 2D and 3Dtunnel models with frequency domain full waveform inversion[J]. PAMM, 2014, 14: 781-782.

[83] VENSTAD J. Industry-scale finite-difference elastic wave modeling on graphics processingunits using the out-of-core technique[J]. Geophysics, 2016, 81: T29-T37.

[84] KREBS J R, ANDERSON J E, HINKLEY D, et al. Fast full-wavefield seismic inversion usingencoded sources[J]. Geophysics, 2009, 74(6): WCC177-WCC188.

[85] GUITTON A, DíAZ E. Attenuating crosstalk noise with simultaneous source full waveforminversion[J]. Geophysical Prospecting, 2012, 60(4): 759-768.

[86] 雷涛. 时间域全波形反演方法及最优化理论研究[D]. 中国地质大学, 2019.

[87] RUBIO F, HANZICH M, FARRéS A, et al. Finite-difference staggered grids in GPUs foranisotropic elastic wave propagation simulation[J]. Computers & Geosciences, 2014, 70: 181-189.

[88] ROMERO L A, GHIGLIA D C, OBER C C, et al. Phase encoding of shot records in prestackmigration[J]. Geophysics, 2000, 65(2): 426-436.

[89] GAUTHIER O, VIRIEUX J, TARANTOLA A. Two‐dimensional nonlinear inversion of seismicwaveforms: numerical results[J]. Geophysics, 1986, 51(7): 1387-1403.

[90] ANDERSON J E, TAN L, WANG D. Time-reversal checkpointing methods for RTM and FWI[J]. Geophysics, 2012, 77(4): S93-S103.

[91] 王保利, 高静怀, 陈文超, 等. 地震叠前逆时偏移的有效边界存储策略[J]. 地球物理学报,2012, 55: 2412-2421.

[92] 杨仁虎, 凌云, 瞿立建, 等. 基于边界存储的地震波场重构及逆时偏移成像应用研究[J]. 地球物理学进展, 2017, 32: 1286-1289.

[93] 段沛然, 谷丙洛, 李振春. 基于优化算子边界存储策略的高效逆时偏移方法[J]. 石油地球物理勘探, 2019, 54: 93-101.

[94] SYMES W W. Reverse time migration with optimal checkpointing[J]. Geophysics, 2007, 72(5): SM213-SM221.

[95] CLAPP R G. Reverse time migration with random boundaries[J]. SEG Technical ProgramExpanded Abstracts, 2009: 2809-2813.

[96] ARNULF A, HARDING A, SINGH S, et al. Nature of upper crust beneath the Lucky Strikevolcano using elastic full waveform inversion of streamer data[J]. Geophysical Journal International,2014, 196: 1471-1491.

[97] 张庆臣. 时间域弹性波动方程全波形反演方法研究[D]. 中国石油大学（北京）, 2019.

[98] 梁展源. 非均匀介质弹性参数地震波形反演方法研究[D]. 中国石油大学(华东), 2023.

[99] 李松龄. 复杂衰减介质数值模拟与全波形反演方法研究[D]. 东北石油大学, 2023.

[100] WANG Y, ZHOU H, CHEN H, et al. Adaptive stabilization for Q-compensated reverse timemigration[J]. Geophysics, 2018, 83(1): S15-S32.

[101] ZHAO X, ZHOU H, WANG Y, et al. A stable approach for Q-compensated viscoelastic reversetime migration using excitation amplitude imaging condition[J]. Geophysics, 2018, 83(5): S459-S476.

[102] 梁旭, 马越, 刘超, 等. 基于多目标函数的黏弹性全波形反演[J]. 煤田地质与勘探, 2023, 51:152-163.

[103] ARAGAO O, SAVA P. Elastic full-waveform inversion with probabilistic petrophysical modelconstraints[J]. Geophysics, 2020, 85(2): R101-R111.

[104] SINGH S, TSVANKIN I, NAEINI E Z. Full-waveform inversion with borehole constraints forelastic VTI media[J]. Geophysics, 2020, 85(6): R553-R563.

[105] GREEN P J. Reversible jump Markov chain Monte Carlo computation and Bayesian modeldetermination[J]. Biometrika, 1995, 82(4): 711-732.

[106] MALINVERNO A. Parsimonious Bayesian Markov Chain Monte Carlo inversion in a nonlineargeophysical problem[J]. Geophysical Journal International, 2002, 151: 675-688.

[107] MALINVERNO A, LEANEY W S. Monte-Carlo Bayesian look-ahead inversion of walkawayvertical seismic profiles[J]. Geophysical Prospecting, 2010, 53(5): 689-703.

[108] SUN J, INNANEN K, ZHANG T, et al. Implicit seismic full waveform inversion withdeep neural representation[J]. Journal of Geophysical Research: Solid Earth, 2023, 128(3):e2022JB025964.

[109] 国运东. 基于多震源和稀疏变换的地震波形反演方法研究[D]. 中国石油大学(华东),2023.

[110] SONG Z, WILLIAMSON P R, PRATT R G. Frequency‐domain acoustic‐wave modeling andinversion of crosshole data: part II—Inversion method, synthetic experiments and real‐dataresults[J]. Geophysics, 1995, 60(3): 796-809.

[111] CHOI Y, ALKHALIFAH T. Source-independent time-domain waveform inversion using convolvedwavefields: application to the encoded multisource waveform inversion[J]. Geophysics,2011, 76(5): R125-R134.

[112] FICHTNER A. Full seismic waveform modelling and inversion[M]. Springer, 2011.

[113] ZHANG Q J, DAI S K, CHEN L, et al. Two-dimensional frequency-domain acoustic fullwaveforminversion with rugged topography[J]. Applied Geophysics, 2015, 12: 378-388.

[114] SUN Y C, ZHANG W, XU J, et al. Numerical simulation of 2D seismic wave propagation inthe presence of a topographic fluid-solid interface at the sea bottom by the curvilinear grid finitedifference method[J]. Geophysical Journal International, 2017, 210: 1721–1738.

[115] 何兵红. 基于无网格的波动方程地震正反演方法研究[D]. 中国石油大学(华东), 2018.

[116] 岳晓鹏. 全波形反演方法技术研究[D]. 长安大学, 2019.

[117] 曹岑. 地震波谱元法正演及地质信息复合约束多参数全波形反演[D]. 中南大学, 2022.

[118] ZHANG W, ZHANG Z, CHEN X. Three-dimensional elastic wave numerical modelling inthe presence of surface topography by a collocated-grid finite-difference method on curvilineargrids[J]. Geophysical Journal International, 2012, 190(1): 358-378.

[119] ZHANG C, ZHANG W. Efficient two-dimensional acoustic wave finite-difference numericalsimulation in strongly heterogeneous media using the adaptive mesh refinement (AMR) technique[J]. Geophysics, 2021, 87: 1-76.

[120] 王艺豪. 近地表地震波速度反演、成像与时间域偏移速度分析方法研究[D]. 中国科学技术大学, 2021.

[121] ZHANG W, CHEN X. Traction image method for irregular free surface boundaries in finitedifference seismic wave simulation[J]. Geophysical Journal International, 2006, 167(1): 337-353.

[122] EAID M V, KEATING S D, INNANEN K A. Multiparameter seismic elastic full-waveforminversion with combined geophone and shaped fiber-optic cable data[J]. Geophysics, 2020, 85(6): R537-R552.

[123] YANG J, ZHU H, LI X, et al. Estimating P wave velocity and attenuation structures using fullwaveform inversion based on a time domain complex-valued viscoacoustic wave equation: themethod[J]. Journal of Geophysical Research: Solid Earth, 2020, 125(6): e2019JB019129.

[124] LIU T, BOHLEN T. Time-domain poroelastic full-waveform inversion of shallow seismicdata: methodology and sensitivity analysis[J]. Geophysical Journal International, 2022, 232(3): 1803-1820.

[125] 唐杰, 刘英昌, 韩盛元, 等. 黏弹性VTI 介质微地震正演模拟与震源机制全波形反演[J]. 地球物理学报, 2022, 65(1): 301-319.

[126] GAO L, PAN Y, RIEDER A, et al. Multiparameter 2-D viscoelastic full-waveform inversionof Rayleigh waves: a field experiment at Krauthausen test site[J]. Geophysical Journal International,2023, 234(1): 297-312.

[127] HE W, LI Y, LIU L. Elastic vertically transversely isotropic full-waveform inversion: fromsynthetic to field data[J]. Geophysics, 2023, 88(3): R385-R405.

[128] LI Q, WU G, JIA Z, et al. Full-waveform inversion in acoustic-elastic coupled media withirregular seafloor based on the generalized finite-difference method[J]. Geophysics, 2023, 88(2): T83-T100.

[129] XING G, ZHU T. Compensating attenuation effects in full-waveform inversion with dissipationdispersiondecoupling[J]. Geophysics, 2023, 88(5): R645-R654.

[130] XIE Y, RYCHERT C A, HARMON N. Elastic and anelastic adjoint tomography with and fullHessian kernels[J]. Geophysical Journal International, 2023, 234(2): 1205-1235.

[131] RASHT-BEHESHT M, HUBER C, SHUKLA K, et al. Physics-informed neural networks(PINNs) for wave propagation and full waveform inversions[J]. Journal of Geophysical Research:Solid Earth, 2022, 127(5): e2021JB023120.

[132] WU Y, MCMECHAN G A, WANG Y. Adaptive feedback convolutional-neural-network-basedhigh-resolution reflection-waveform inversion[J]. Journal of Geophysical Research: SolidEarth, 2022, 127(6): e2022JB024138.

[133] 许屹然. 智能化声波方程正反演理论与方法[D]. 中国石油大学(北京), 2022.

[134] LI D, GUO Y, PENG S, et al. Seismic full-waveform inversion based on superresolution forhigh-precision prediction of reservoir parameters[J]. Geophysics, 2023, 88(3): R243-R256.

[135] YANG F, MA J. FWIGAN: full-waveform inversion via a physics-informed generativeadversarial network[J]. Journal of Geophysical Research: Solid Earth, 2023, 128(4):e2022JB025493.

[136] YAO J, WARNER M, WANG Y. Regularization of anisotropic full-waveform inversion withmultiple parameters by adversarial neural networks[J]. Geophysics, 2023, 88(1): R95-R103.

[137] LIU Q, PETER D. Square-root variable metric-based nullspace shuttle: a characterization of thenonuniqueness in elastic full-waveform inversion[J]. Journal of Geophysical Research: SolidEarth, 2020, 125(2): e2019JB018687.

[138] WANG T, CHENG J, GENG J. Reflection full waveform inversion with second-order optimizationusing the adjoint-state method[J]. Journal of Geophysical Research: Solid Earth, 2021, 126(8): e2021JB022135.

[139] YONG P, BROSSIER R, MéTIVIER L. Parsimonious truncated Newton method for timedomainfull-waveform inversion based on the Fourier-domain full-scattered-field approximation[J]. Geophysics, 2022, 87(1): R123-R146.

[140] KEATING S D, INNANEN K A. Regularization tunnelling for full waveform inversion[J].Geophysical Journal International, 2022, 229(3): 1503-1516.

[141] 刘宇航, 黄建平, 杨继东, 等. 弹性波全波形反演中的四种优化方法对比[J]. 石油地球物理勘探, 2022, 57: 118-128.

[142] GHOLAMI A, AGHAMIRY H S, OPERTO S. Multiplier waveform inversion: a reduced-spacefull-waveform inversion by the method of multipliers[J]. Geophysics, 2023, 88(3): R339-R354.

[143] CARVALHO P T C, DA SILVA S L E F, DUARTE E F, et al. Full waveform inversion basedon the non-parametric estimate of the probability distribution of the residuals[J]. GeophysicalJournal International, 2021, 229(1): 35-55.

[144] GHOLAMI A, AGHAMIRY H S, OPERTO S. Extended-space full-waveform inversion in thetime domain with the augmented Lagrangian method[J]. Geophysics, 2022, 87(1): R63-R77.

[145] LI D, LAMOUREUX M P, LIAO W. Application of an unbalanced optimal transport distanceand a mixed L1/Wasserstein distance to full waveform inversion[J]. Geophysical Journal International,2022, 230(2): 1338-1357.

[146] OPERTO S, GHOLAMI A, AGHAMIRY H S, et al. Full waveform inversion beyond the Bornapproximation: a tutorial review[A]. 2023: 2212.10141.

[147] ZHANG H, HE W, MA J. The back-and-forth method for the quadratic Wasserstein distancebasedfull-waveform inversion[J]. Geophysics, 2023, 88(4): R469-R483.

[148] HUANG X, EIKREM K S, JAKOBSEN M, et al. Bayesian full-waveform inversion inanisotropic elastic media using the iterated extended Kalman filter[J]. Geophysics, 2020, 85(4): C125-C139.

[149] ZHANG X, CURTIS A. Bayesian full-waveform inversion with realistic priors[J]. Geophysics,2021, 86(5): A45-A49.

[150] KAN L Y, CHEVROT S, MONTEILLER V. A consistent multiparameter Bayesian full waveforminversion scheme for imaging heterogeneous isotropic elastic media[J]. Geophysical JournalInternational, 2022, 232(2): 864-883.

[151] YE W, HUANG X, HAN L, et al. Bayesian inverse scattering theory for multiparameter fullwaveforminversion in transversely isotropic media[J]. Geophysics, 2023, 88(3): C91-C109.

[152] ZHANG X, LOMAS A, ZHOU M, et al. 3-D Bayesian variational full waveform inversion[J].Geophysical Journal International, 2023, 234(1): 546-561.

[153] GóRSZCZYK A, BROSSIER R, MéTIVIER L. Graph-space optimal transport concept fortime-domain full-waveform inversion of ocean-bottom seismometer data: nankai trough velocitystructure reconstructed from a 1D model[J]. Journal of Geophysical Research: Solid Earth,2021, 126(5): e2020JB021504.

[154] LIU X, ZHU T, HAYES J. Critical zone structure by elastic full waveform inversion of seismicrefractions in a sandstone catchment, central pennsylvania, USA[J]. Journal of GeophysicalResearch: Solid Earth, 2022, 127(3): e2021JB023321.

[155] MA J, BUNGE H P, THRASTARSON S, et al. Seismic Full-Waveform Inversion of the Crust-Mantle Structure Beneath China and Adjacent Regions[J]. Journal of Geophysical Research:Solid Earth, 2022, 127(9): e2022JB024957.

[156] ZHOU T, LI J, XI Z, et al. CUSRA2021: a radially anisotropic model of the contiguous US andsurrounding regions by full-waveform inversion[J]. Journal of Geophysical Research: SolidEarth, 2022, 127(8): e2021JB023893.

[157] KAN L Y, CHEVROT S, MONTEILLER V. Dehydration of the subducting juan de fuca plateand fluid pathways revealed by full waveform inversion of teleseismic P and SH waves in centraloregon[J]. Journal of Geophysical Research: Solid Earth, 2023, 128(4): e2022JB025506.

[158] ZHOU W, LUMLEY D. Nonrepeatability effects on time-lapse 4D seismic full-waveform inversionfor ocean-bottom node data[J]. Geophysics, 2021, 86(4): R547-R561.

[159] LIU Y, TSVANKIN I. Source-independent time-lapse full-waveform inversion for anisotropicmedia[J]. Geophysics, 2022, 87(1): R111-R122.

[160] FU X, INNANEN K A. Stepsize sharing in time-lapse full-waveform inversion[J]. Geophysics,2023, 88(2): M59-M70.

[161] HUANG C, ZHU T, XING G. Data-assimilated time-lapse visco-acoustic full-waveform inversion:theory and application for injected CO2 plume monitoring[J]. Geophysics, 2023, 88(1): R105-R120.

[162] LIU X, ZHU T, AJO-FRANKLIN J. Understanding subsurface fracture evolution dynamicsusing time-lapse full waveform inversion of continuous active-source seismic monitoring data[J]. Geophysical Research Letters, 2023, 50(4): e2022GL101739.

[163] WANG H, GUO Q, ALKHALIFAH T, et al. Regularized elastic passive equivalent sourceinversion with full-waveform inversion: application to a field monitoring microseismic data set[J]. Geophysics, 2020, 85(6): KS207-KS219.

[164] AGHAMIRY H S, GHOLAMI A, OPERTO S, et al. ADMM-based full-waveform inversionfor microseismic imaging[J]. Geophysical Journal International, 2021, 228(1): 259-274.

[165] XU Z, MIKESELL T D. Estimation of resolution and covariance of ambient seismic sourcedistributions: full waveform inversion and matched field processing[J]. Journal of GeophysicalResearch: Solid Earth, 2022, 127(6): e2022JB024374.

[166] ROBERTSSON J O A, ANDERSSON F, ÉDOUARD PLESSIX R. Efficient snapshot-freereverse time migration and computation of multiparameter gradients in full-waveform inversion[J]. Geophysics, 2021, 86(5): T305-T320.

[167] THRASTARSON S, VAN HERWAARDEN D P, KRISCHER L, et al. Data-adaptive globalfull-waveform inversion[J]. Geophysical Journal International, 2022, 230(2): 1374-1393.

[168] 齐红宇, 傅红, 笋杨露. 应用修正正交有限内存拟牛顿算法的全波形反演[J]. 石油地球物理勘探, 2022, 57(5): 1114-1119.

[169] 秦宁, 梁鸿贤, 郭振波, 等. 基于区域分解的3D 弹性波全波形反演并行实现策略[J]. 石油地球物理勘探, 2023, 58(2): 351-357.

[170] XU M, ZHAO X, CANALES J P. Structural variability within the kane oceanic core complexfrom full waveform inversion and reverse time migration of streamer data[J]. GeophysicalResearch Letters, 2020, 47(7): e2020GL087405.

[171] CAO J, BROSSIER R, GóRSZCZYK A, et al. 3-D multiparameter full-waveform inversionfor ocean-bottom seismic data using an efficient fluid–solid coupled spectral-element solver[J].Geophysical Journal International, 2021, 229(1): 671-703.

[172] KAMATH N, BROSSIER R, MéTIVIER L, et al. Multiparameter full-waveform inversion of3D ocean-bottom cable data from the Valhall field[J]. Geophysics, 2021, 86(1): B15-B35.

[173] PLADYS A, BROSSIER R, KAMATH N, et al. Robust full-waveform inversion with graphspaceoptimal transport: application to 3D ocean-bottom cable Valhall data[J]. Geophysics,2022, 87(3): R261-R280.

[174] BORISOV D, GAO F, WILLIAMSON P, et al. Application of 2D full-waveform inversion onexploration land data[J]. Geophysics, 2020, 85(2): R75-R86.

[175] RETA-TANG C, SHENG J, LIU F, et al. Application of full-waveform inversion to land data:case studies in onshore Mexico[J]. The Leading Edge, 2023, 42(3): 190-195.

[176] SINGH S, SEARS T, ROBERTS M, et al. Full elastic waveform inversion: future of quantitativeseismic imaging[J]. SEG Technical Program Expanded Abstracts, 2008, 27: 1905-1909.

[177] CRASE E, WIDEMAN C, NOBLE M, et al. Nonlinear elastic waveform inversion of landseismic reflection data[J]. Journal of Geophysical Research: Solid Earth, 1992, 97(B4): 4685-4703.

[178] BAUMSTEIN A, ANDERSON J E, HINKLEY D L, et al. Scaling of the objective functiongradient for full wavefield inversion[J]. SEG Technical Program Expanded Abstracts, 2009:2243-2247.

[179] VIGH D, JIAO K, WATTS D, et al. Elastic full-waveform inversion application using multicomponentmeasurements of seismic data collection[J]. Geophysics, 2014, 79: R63–R77.

[180] REN Z, LIU Y. A hierarchical elastic full-waveform inversion scheme based on wavefieldseparation and the multistep-length approach[J]. Geophysics, 2016, 81: R99-R123.

[181] JEONG W, LEE H Y, MIN D J. Full waveform inversion strategy for density in the frequencydomain[J]. Geophysical Journal International, 2012, 188(3): 1221-1242.

[182] WANG T, CHENG J. Elastic full waveform inversion based on mode decomposition: theapproach and mechanism[J]. Geophysical Journal International, 2017, 209: 606-622.

[183] WEISS R M, SHRAGGE J. Solving 3D anisotropic elastic wave equations on parallel GPUdevices[J]. Geophysics, 2013, 78(2): F7-F15.

[184] JIANG J, ZHU P. Acceleration for 2D time-domain elastic full waveform inversion using asingle GPU card[J]. Journal of Applied Geophysics, 2018, 152: 173-187.

[185] LI Y, ALKHALIFAH T. Target-oriented high-resolution elastic full-waveform inversion withan elastic redatuming method[J]. Geophysics, 2022, 87(5): R379-R389.

[186] FANG J, ZHOU H, ZHANG Q, et al. The effects of elastic data on acoustic and elastic fullwaveform inversion[J]. Journal of Applied Geophysics, 2019, 172: 103876.

[187] SEARS T, SINGH S, BARTON P. Elastic full waveform inversion of multi-component OBCseismic data[J]. Geophysical Prospecting, 2008, 56: 843-862.

[188] SEARS T J, BARTON P J, SINGH S C. Elastic full waveform inversion of multicomponentocean-bottom cable seismic data: application to Alba Field, U. K. North Sea[J]. Geophysics,2010, 75(6): R109-R119.

[189] KöHN D, DE NIL D, KURZMANN A, et al. On the influence of model parametrization inelastic full waveform tomography[J]. Geophysical Journal International, 2012, 191: 325-345.

[190] RAKNES E, ARNTSEN B, WEIBULL W. Three-dimensional elastic full waveform inversionusing seismic data from the Sleipner area[J]. Geophysical Journal International, 2015, 202:1877-1894.

[191] HUSTEDT B, OPERTO S, VIRIEUX J. Mixed-grid and staggered-grid finite-difference methodsfor frequency-domain acoustic wave modelling[J]. Geophysical Journal International, 2004,157(3): 1269-1296.

[192] THIEL N, HERTWECK T, BOHLEN T. Comparison of acoustic and elastic full‐waveforminversion of 2D towed‐streamer data in the presence of salt[J]. Geophysical Prospecting, 2018,67: 349–361.

[193] BRENDERS A J, PRATT R G. Full waveform tomography for lithospheric imaging: resultsfrom a blind test in a realistic crustal model[J]. Geophysical Journal International, 2007, 168(1): 133-151.

[194] BARNES C, CHARARA M. Full-waveform inversion results when using acoustic approximationinstead of elastic medium[J]. SEG Technical Program Expanded Abstracts, 2008, 27:1895-1899.

[195] BARNES C, CHARARA M. The domain of applicability of acoustic full-waveform inversionfor marine seismic data[J]. Geophysics, 2009, 74: WCC105-WCC108.

[196] AGUDO O C, DA SILVA N V, WARNER M, et al. Acoustic full-waveform inversion in anelastic world[J]. Geophysics, 2018, 83(3): R257-R271.

[197] MARELLI S, MAURER H, MANUKYAN E. Validity of the acoustic approximation in fullwaveformseismic crosshole tomography[J]. Geophysics, 2012, 77: R129-R139.

[198] HOBRO J, CHAPMAN C, ROBERTSSON J. A method for correcting acoustic finite-differenceamplitudes for elastic effects[J]. Geophysics, 2014, 79: T243-T255.

[199] MULDER W, PLESSIX R E. Exploring some issues in acoustic full waveform inversion[J].Geophysical Prospecting, 2008, 56: 827 - 841.

[200] CHAPMAN C, HOBRO J, ROBERTSSON J. Elastic corrections to acoustic finite‐differencesimulations[J]. SEG Technical Program Expanded Abstracts, 2010: 3013-3017.

[201] CHAPMAN C, HOBRO J, ROBERTSSON J. Correcting an acoustic wavefield for elasticeffects[J]. Geophysical Journal International, 2014, 197: 1196-1214.

[202] LI Q, WU G. 2D multi-parameter waveform inversion of land reflection seismic data obtainedfrom the particle-motion response from the vertical geophone[J]. Acta Geophysica, 2020, 68:377-388.

[203] 李青阳, 吴国忱, 王玉梅, 等. 基于最优输运原理的陆上单分量资料弹性波全波形反演[J].石油地球物理勘探, 2021, 56(5): 1060-1073.

[204] BIONDI B, ALMOMIN A. Tomographic full-waveform inversion (TFWI) by combining FWIand wave-equation migration velocity analysis[J]. The Leading Edge, 2013, 32(9): 1074-1080.

[205] BAETEN G, MAAG J, PLESSIX R E, et al. The use of low frequencies in a full‐waveforminversion and impedance inversion land seismic case study[J]. Geophysical Prospecting, 2013,61: 701-711.

[206] 管路平, 唐权钧. 地震信号的高低频成分补偿[J]. 石油物探, 1990, 29(3): 35-45.

[207] WARNER M, NANGOO T, SHAH N, et al. Full-waveform inversion of cycle-skipped seismicdata by frequency down-shifting[M]//SEG Technical Program Expanded Abstracts 2013. 2013:903-907.

[208] HU W. FWI without low frequency data - beat tone inversion[J]. SEG Technical ProgramExpanded Abstracts, 2014: 1116-1120.

[209] SUN B, ALKHALIFAH T. An anisotropic waveform inversion using an optimal transportmatching filter objective: an application to an offshore field dataset[J]. IEEE Transactions onGeoscience and Remote Sensing, 2023, 61: 1-22.

[210] HU W, JIN Y, WU X, et al. Physics-guided self-supervised learning for low frequency dataprediction in FWI[J]. SEG Technical Program Expanded Abstracts, 2020: 875-879.

[211] GOODFELLOW I J, POUGET-ABADIE J, MIRZA M, et al. Generative adversarial nets[J].MIT Press, 2014: 2672–2680.

[212] ZHU J Y, PARK T, ISOLA P, et al. Unpaired Image-to-Image Translation Using Cycle-Consistent Adversarial Networks[J]. 2017 IEEE International Conference on Computer Vision(ICCV), 2017: 2242-2251.

[213] MARFURT K J. Accuracy of finite-difference and finite-element modeling of the scalar andelastic wave equations[J]. Geophysics, 1984, 49(5): 533-549.

[214] SERIANI G, PRIOLO E. Spectral element method for acoustic wave simulation in heterogeneousmedia[J]. Finite Elements in Analysis and Design, 1994, 16(3): 337-348.

[215] KOMATITSCH D, BARNES C, TROMP J. Simulation of anisotropic wave propagation basedupon a spectral element method[J]. Geophysics, 2000, 65(4): 1251-1260.

[216] KäSER M, DUMBSER M. An arbitrary high-order discontinuous Galerkin method for elasticwaves on unstructured meshes –I. The two-dimensional isotropic case with external sourceterms[J]. Geophysical Journal International, 2006, 166(2): 855-877.

[217] VIRIEUX J. SH-wave propagation in heterogeneous media: velocity‐stress finite‐differencemethod[J]. Geophysics, 1984, 49(11): 1933-1942.

[218] VIRIEUX J. P-SV wave propagation in heterogeneous media: velocity‐stress finite‐differencemethod[J]. Geophysics, 1986, 51(4): 889-901.

[219] CLAYTON R, ENGQUIST B. Absorbing boundary conditions for acoustic and elastic waveequations[J]. Bulletin of the Seismological Society of America, 1977, 67(6): 1529-1540.

[220] CERJAN C, KOSLOFF D, KOSLOFF R, et al. A nonreflecting boundary condition for discreteacoustic and elastic wave equations[J]. Geophysics, 1985, 50(4): 705-708.

[221] BERENGER J P. A perfectly matched layer for the absorption of electromagnetic waves[J].Journal of Computational Physics, 1994, 114(2): 185-200.

[222] KOMATITSCH D, MARTIN R. An unsplit convolutional perfectly matched layer improved atgrazing incidence for the seismic wave equation[J]. Geophysics, 2007, 72(5): SM155-SM167.

[223] KINCAID D R, CHENEY E W. Numerical analysis[M]. mathematics of scientific computing,2002.

[224] MODRAK R, TROMP J. Seismic waveform inversion best practices: regional, global andexploration test cases[J]. Geophysical Journal International, 2016, 206(3): 1864-1889.

[225] GARDNER G, GARDNER L, GREGORY A. Formation velocity and density—the diagnosticbasics for stratigraphic traps[J]. Geophysics, 1974, 39: 770-780.

[226] MALINOWSKI M, OPERTO S, RIBODETTI A. High-resolution seismic attenuation imagingfrom wide-aperture onshore data by visco-acoustic frequency-domain full-waveform inversion[J]. Geophysical Journal International, 2011, 186(3): 1179-1204.

[227] MARTIN G, WILEY R, MARFURT K. Marmousi2: an elastic upgrade for Marmousi[J]. TheLeading Edge, 2006, 25: 156-166.

[228] WEIBULL W W, ARNTSEN B. Automatic velocity analysis with reverse-time migration[J].Geophysics, 2013, 78(4): S179-S192.

[229] MEHRALIAN M, KARASFI B. RDCGAN: unsupervised representation learning with regularizeddeep convolutional generative adversarial networks[J]. Conference on Artificial Intelligenceand Robotics and 2nd Asia-Pacific International Symposium, 2018: 31-38.

[230] KARRAS T, AILA T, LAINE S, et al. Progressive growing of GANs for improved quality,stability, and variation[A]. 2017: abs/1710.10196.

[231] SALIMANS T, GOODFELLOW I, ZAREMBA W, et al. Improved techniques for trainingGANs[J]. Curran Associates Inc, 2016: 2234–2242.

[232] ARJOVSKY M, CHINTALA S, BOTTOU L. Wasserstein generative adversarial networks[J].Proceedings of the 34th International Conference on Machine Learning - Volume 70, 2017:214–223.

[233] ISOLA P, ZHU J Y, ZHOU T, et al. Image-to-image translation with conditional adversarialnetworks[J]. 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR),2017: 5967-5976.

[234] SUN Y, ZHANG W, CHEN X. 3D seismic wavefield modeling in generally anisotropic mediawith a topographic free surface by the curvilinear grid finite‐difference method[J]. Bulletin ofthe Seismological Society of America, 2018, 108(3A): 1287-1301.

[235] MORA P. Nonlinear two‐dimensional elastic inversion of multioffset seismic data[J]. Geophysics,1987, 52(9): 1211-1228.

[236] KJARTANSSON E. Constant Q-wave propagation and attenuation[J]. Journal of GeophysicalResearch: Solid Earth, 1979, 84(B9): 4737-4748.

[237] WILSON M P, DAVIES R J, FOULGER G R, et al. Anthropogenic earthquakes in the UK: anational baseline prior to shale exploitation[J]. Marine and Petroleum Geology, 2015, 68: 1-17.

[238] TURNER C C, CRONIN B T. Rift-related coarse-grained submarine fan reservoirs; the braeplay, south viking graben, north Sea[M]. The American Association of Petroleum Geologists,2018.

[239] KEYS R G, FOSTER D J. A data set for evaluating and comparing seismic inversion methods[M]. Society of Exploration Geophysicists, 1998: 1-12.

[240] MADIBA G B, MCMECHAN G A. Processing, inversion, and interpretation of a 2D seismicdata set from the North Viking Graben, North Sea[J]. Geophysics, 2003, 68(3): 837-848.

[241] GUO Q, ALKHALIFAH T. Target-oriented waveform redatuming and high-resolution inversion:role of the overburden[J]. Geophysics, 2020, 85(6): R525-R536.

[242] GUO P, SINGH S C, VADDINENI V A, et al. Nonlinear full waveform inversion of wideapertureOBS data for Moho structure using a trans-dimensional Bayesian method[J]. GeophysicalJournal International, 2020, 224(2): 1056-1078.

[243] DONG ZHANG Z, ALKHALIFAH T. High-resolution reservoir characterization using deeplearning-aided elastic full-waveform inversion: the north sea field data example[J]. Geophysics,2020, 85(4): WA137-WA146.

[244] KIM J, KIM M, KANG H, et al. U-GAT-IT: unsupervised generative attentional networkswith adaptive layer-instance normalization for image-to-image translation[A]. 2019:abs/1907.10830.

[245] TANG H, LIU H, XU D, et al. AttentionGAN: unpaired image-to-image translation usingattention-guided generative adversarial networks[J]. IEEE Transactions on Neural Networksand Learning Systems, 2023, 34(4): 1972-1987.